Selection for Enhanced Disease Resistance and Growth Performance in Cross-bred Eastern Oysters, Crassostrea virginica

Transcription

1 The University of Maine University of Maine Office of Research and Sponsored Programs: Grant Reports Special Collections Selection for Enhanced Disease Resistance and Growth Performance in Cross-bred Eastern Oysters, Crassostrea virginica Paul Rawson Principal Investigator; University of Maine, Orono, Follow this and additional works at: Part of the Aquaculture and Fisheries Commons Recommended Citation Rawson, Paul, "Selection for Enhanced Disease Resistance and Growth Performance in Cross-bred Eastern Oysters, Crassostrea virginica" (2014). University of Maine Office of Research and Sponsored Programs: Grant Reports. Paper This Open-Access Report is brought to you for free and open access by It has been accepted for inclusion in University of Maine Office of Research and Sponsored Programs: Grant Reports by an authorized administrator of

2 Project Completion Report Project Title: Selection for enhanced disease resistance and growth performance in cross-bred eastern oysters, Crassostrea virginica. Subaward # Z Grant # SIGNATURE PAGE PROJECT CODE: TRA & TRA SUBCONTRACT NO: Z PROJECT TITLE: Selection for enhanced disease resistance and growth performance in cross-bred eastern oysters, Crassostrea virginica. PREPARED BY: Dr. Paul Rawson School of Marine Sciences University of Maine Orono, ME Project Coordinator Date 1

3 Project Completion Report Project Title: Selection for enhanced disease resistance and growth performance in cross-bred eastern oysters, Crassostrea virginica. Subaward # Z Grant # PROJECT CODE: TRA & TRA SUBCONTRACT/ACCOUNT NO: Z PROJECT TITLE: Selection for enhanced disease resistance and growth performance in cross-bred eastern oysters, Crassostrea virginica. DATES OF WORK: July 1, 2009 to June 30, 2013 PARTICIPANTS: Paul Rawson, University of Maine Scott Lindell, Marine Biological Laboratory Ximing Guo, Rutgers University Inke Sunila, CT Bureau of Aquaculture Chris Davis, Pemaquid Oyster Co. Dana Morse, Maine Sea Grant Diane Murphy, Cape Cod Cooperative Extension REASON FOR TERMINATION: Project completed. PROJECT OBJECTIVES: 1) Through a replicated field trial we will test the hypothesis that enhanced disease resistance to multiple diseases and improved growth can be realized through a combination of interline crossing and genetic selection on cross-bred lines o f oysters. 2) Together with oyster growers, hatchery operators and extension agents, we will develop guidelines and protocols for the transfer of the best performing lines from our program to commercial hatcheries and the industry. Project results and updates on access to broodstock will be presented at the annual NACE meeting, at annual meetings of the National Shellfisheries Association and via a Fact Sheet at the project s end. ANTICIPATED BENEFITS: The Northeast region has the largest oyster culture industry on the Atlantic coast. Yet, outbreaks of four diseases, MSX, SSO, Dermo, and ROD, cause considerable damage to the industry and limit expansion. Despite considerable effort from both industry and academia, a single line that resists all four diseases and grows well does not currently exist. This project sought to identify lines of oysters that perform well at a variety of sites in the region, particularly those sites where oyster diseases are endemic, by virtue of developing lines of oysters that are resistant to multiple diseases. As such, the major product generated by this project is a repository of lines that have demonstrated high performance, including disease resistance, at sites in New England. In addition, the results o f our 2

4 field trial will help growers choose seed given their particular grow-out site s characteristics (e.g., their anticipated disease exposure). Dissemination of the results and key information from this project is ongoing and our extension efforts will take place on three levels. First, the most direct approach is through direct contact with growers. A particular strength of our project was the direct involvement of commercial growers. Several of the PIs involved in this project have also met regularly with industry working groups, such as the Maine Oyster Growers Working Group, East Coast Shellfish Growers, and individual state aquaculture associations to present updates on the project. We are continuing such efforts even though the funding for this project has terminated. In these efforts, we gained considerable assistance from extension personnel in each state. Second, we will use electronic media to post and store information gained from the project (central website in development, should be live by September 2014). We plan to expand such efforts through additional communications, including dissemination of this final report along with a brief summary of the results from both this project and its predecessor (NRAC #Q169601/Q169901) via and posting on the website of the East Coast Shellfish Growers Association (ECSGA). Third, we will submit formal journal manuscripts on the production and disease-resistant characteristics of the lines used in our project. In addition, we are planning to convene a meeting of interested growers and hatchery operators at the NACE 2015 meeting in Portland, to review project results, the availability of lines and future research plans. In the meantime, the DMC hatchery is propagating the Clinton and UMFS lines while the Haskins Shellfish lab continues to maintain the NEH lines and make these available to hatcheries and growers in the region. PRINCIPAL ACCOMPLISHMENTS: Oyster seed from six lines were deployed at grow-out sites from New Jersey to Maine in late July of The University of Maine hatchery provided seed for three replicate grow-out bags of the UMFS, Clinton, UMFS x Clinton, UMFS x NEH (new F 1), and UMFS x NEH F2 lines while triploid seed from the NEH line was purchased from a commercial hatchery. These lines were deployed at six industry partner sites, including two in Maine by Rawson (University of Maine), one in each of Massachusetts and Rhode Island by Lindell (MBL), one in Connecticut by Sunila (CT Bureau of Aquaculture) and at the Haskin Shellfish Laboratory s Cape Shore site by Guo. The performance (mortality and size) of oysters in all deployed bags was assessed every 2-3 months during two growing seasons (2011 and 2012). During the fall of 2011, we monitored oysters deployed at sites in Maine and Connecticut for prevalence of Roseovarius Oyster Disease (ROD). In addition, 30 oysters from each line grown at each site were sampled for histopathology-based assessment of disease pressure from MSX, SSO, and Dermo. At the conclusion of the field trial in the late fall of 2012, oysters in all replicates were harvested, and final measurements of survival (counts of live and dead oysters), shell height, shell width, and wet weight were taken and yield for each replicate bag estimated. The accompanying final technical report provides the details regarding our assessment of relative performance (disease resistance, growth and yield) of each of the three parental and three cross-bred lines included in our trial. Briefly, we found that all three of the parental lines used in our study exhibit resistance to MSX. At the same time, we observed line-specific susceptibility to MSX/SSO co-infections, particularly in advanced generation hybrid lines, differential susceptibility to Dermo and to pathological interactions with ciliates when lines were grown under sub-optimal conditions. The line-specific variation in disease susceptibility led to line-specific differences in yield and a 3

5 strong site-specific performance. Such site-specific performance is a clear indication that the parental lines in our study are locally-adapted. We also found that line crossing can provide important performance gains. The intermediate resistance to Dermo combined with high resistance to MSX/SSO co-infections and ROD in the UMFS x NEH F1 indicates that interline crosses will provide growers some measure of protection at sites where disease pressure is mixed, variable or unpredictable. At the same time, the long-term propagation of a hybrid line without continued selection pressure is unlikely to be advantageous, as segregation and independent assortment of parental allele combinations in advanced hybrids appears to result in declines in MSX/SSO resistance. Though we cannot easily explain the shift in the performance of the Clinton line in our study relative to previous studies, we recommend the continued propagation, development, and use of all three parental lines used in our study. We were successful at meeting nearly all of the goals with this project with the exception of performing size-based selection on the UMFS x NEH F 1 line. Due to problems encountered at the start of our project we were unable to obtain and spawn enough of this line to support size-based selection. IMPACTS: The research completed under this award will have three main impacts: 1) Our results provide the northeastern oyster culture industry with improved knowledge base regarding the performance of genetically improved lines and the utility of cross-breeding among lines. Our key findings are as follows: The three lines deployed in our study, NEH, Clinton, and UMFS, exhibit a high degree of local adaptation. The NEH line remains the line of choice in southern New England due to low mortality in SNE sites associated with Dermo resistance combined with fast growth. The UMFS line and hybrid crosses between lines exhibited the highest survival as a result of improved resistance to Roseovarius Oyster Disease (ROD) and better performance at sites where there are co-infections of MSX/SSO. Such conditions are common in northern New England. All three lines, NEH, UMFS, and Clinton, and their hybrids expressed resistance to MSX. Site-specific variation in growth typically outweighs line-specific differences in growth so that variation in yield among these lines is driven by line-specific variation in mortality which, in turn, is highly dependent on disease resistance. Site-specific variation in shell shape and condition index dwarfed any line-specific variation in these features. Lines which are grown at sites that are sub-optimal are prone to pathological lesions due to ciliates and other normally commensal organisms. Cross-bred oysters, particularly those produced by crossing the UMFS and NEH lines, exhibited survival and yield that was intermediate to the parental lines in southern New England and equal to the best parental line (UMFS) in northern New England. Overall, 4

6 the UMFS x NEH hybrid oysters had marginal yields across all six sites in our study equal to the marginal yield of the best parental line, NEH. Intermediate resistance to Dermo combined with high resistance to MSX/SSO coinfections and ROD in the UMFS x NEH F1 line indicates that interline crosses will provide growers some measure of protection at sites where disease pressure is mixed, variable or unpredictable. Even so, the long-term propagation of a hybrid line without continued selection pressure is unlikely to be advantageous. 2) Our results provide an opportunity to map disease resistance of the currently available lines to anticipated disease pressure allowing growers to reduce losses from disease outbreaks. 3) We are continuing to propagate the parental lines used in this study; they are available to commercial hatcheries on request. RECOMMENDED FOLLOW-UP ACTIVITIES: 1) Research on the impact of MSX/SSO co-infections and the nature of resistance to SSO will be critical as the range, prevalence and intensity of SSO increases over time. All lines deployed in our current study expressed resistance to MSX, by itself, and we observed low prevalences and intensities of MSX. However, we also observed line-specific responses to MSX/SSO co-infections and lines developed in southern New England are unlikely to have experienced SSO pressure. Increased attention to SSO will help reduce the likelihood of SSO outbreaks that could have a devastating impact on oyster operations in the Northeast. 2) We have observed substantial line-specific variation in susceptibility to the development of pathological interactions with normally commensal organisms. In particular, we found increased prevalence of lesions associated with ciliates inhabiting the gills of oysters in UMFS oysters cultured at the Cape Shore, NJ site. These observations suggest that the culture environment is suboptimal for UMFS oysters and they are physiologically compromised leading to a reduced ability to control normally commensal organisms. Elucidating the line-specific physiological responses across the range of culture environments in the Northeast region will allow growers to avoid reductions in growth and survival associated with unfavorable physiological interactions. 3) We observed a substantial decline in the performance of the Clinton line in this study compared to our previous field trial. In the previous study, the Clinton line had growth, survival and yield nearly equal to that of the NEH line in southern New England. In the present study, the Clinton line was among the worst of our lines, even at the Clinton site. Research is needed to determine whether inbreeding effects or hatchery-specific conditions underlie the observed decline in the performance of this line. SUPPORT: YEAR NRAC- OTHTER SUPPORT Total USDA University Industry Other Other Total Support Funding Federal 1 $116,622 $25,863 $2,602 $28,465 $145,087 $114,754 $25,937 $5,361 $31,298 $146,052 TOTAL $231,416 $51,800 $7,963 $59,763 $291,139 5

7 PUBLICATIONS, MANUSCRIPTS, OR PAPERS PRESENTED: Oral Presentations Sunila, I., S. Lindell, X. Guo and P. Rawson, Field trials for cross-bred disease-resistant eastern oysters in New England. 32nd Milford Aquaculture Seminar, Westbrook, CT. March J. Shellfish Res. 31 (1): p Rawson, P., X. Guo, S. Lindell, and I. Sunila. Disease-Resistance and Improved Performance for Cross-bred Oysters, Crassostrea virginica: I. Evaluation of Line-Specific Growth and Yield. National Shellfisheries Association 105th Annual Meeting, Nashville, TN. February Sunila, I., S. Lindell, X. Guo, and P. Rawson. Disease-resistance and improved performance for cross-bred eastern oysters Crassostrea virginica. II Evaluation of line-specific disease prevalence. National Shellfisheries Association 105th Annual Meeting, Nashville, TN. February Sunila, I. Keynote Presentation: Development of a MSX resistant oyster strain for industry. Aquatic Aquaculture Association of Nova Scotia. Invasive Species and Disease Workshop. Research and Industry Connector Event. Halifax, Nova Scotia. January, 2014 Rawson, P., S. Lindell, X. Guo, and I. Sunila. Building a Better Oyster for New England. National Shellfisheries Association 106th Annual Meeting, Jacksonville, FL. March Manuscripts (planned) Sunila, I., P. Rawson, S. Lindell and X. Guo History and Current Status of Improved Lines of Eastern Oysters, Crassostrea virginica. NRAC Fact Sheet, in prep. Rawson, P. S. Lindell, I. Sunila, and X. Guo Variation in Line-Specific Survival and Yield among Improved Lines of Eastern Oysters at Northeastern U.S. sites impacted by disease. Aquaculture (planned). Rawson, P. S. Lindell, I. Sunila, and X. Guo Variation in Line-Specific disease presentation among Improved Lines o f Eastern Oysters at Northeastern U.S. sites. Aquaculture (planned). 6

8 Project Final Technical Report Selection for enhanced disease resistance and growth performance in cross-bred eastern oysters, Crassostrea virginica. Project Code: TRA & TRA Subcontract: Z Project Grant Number: Dates of Work: July 1, 2009 to June 30, 2013 Participants Paul Rawson, University o f Maine Scott Lindell, Marine Biological Laboratory Ximing Guo, Rutgers University Inke Sunila, CT Bureau of Aquaculture Chris Davis, Pemaquid Oyster Co. Dana Morse, Maine Sea Grant Diane Murphy, Cape Cod Cooperative Extension Project Objectives 1) Through a replicated field trial we will test the hypothesis that enhanced disease resistance to multiple diseases and improved growth can be realized through a combination of interline crossing and genetic selection on cross-bred lines o f oysters. 2) Together with oyster growers, hatchery operators and extension agents, we will develop guidelines and protocols for the transfer of the best performing lines from our program to commercial hatcheries and the industry. Project results and updates on access to broodstock will be presented at the annual NACE meeting, at annual meetings of the National Shellfisheries Association and via a Fact Sheet at the project s end. 1

9 M e t h o d s a n d P r o c e d u r e s Objective 1 - Through a replicated field trial we will test the hypothesis that enhanced disease resistance to multiple diseases and improved growth can be realized through a combination of interline crossing and genetic selection on cross-bred oysters. Our collaborative broodstock development program seeks to provide commercial and public oyster hatcheries with high performance, disease resistant broodstock so that growers in the Northeast have access to oyster seed with enhanced survival and growth which, in turn, leads to increased oyster production throughout the region. In previous work, we found evidence for hybrid vigor with respect to both growth and disease resistance when two lines of oysters, the University of Maine s UMFS line and Rutgers University s NEH line, were cross-bred. The present project tested whether hybrid vigor can also be realized when the UMFS line is crossed with the disease resistant Clinton line. We also planned to determine whether selection on UMFS-NEH cross-bred oysters can be used to further enhance disease resistance and growth performance. This latter objective was not realized during the project; due to problems encountered at the start of our project we were unable to obtain and spawn enough of this line to support size-based selection. However, we were successful at comparing the performance of the UMFS x Clinton and UMFS x NEH hybrid non-selected oysters to pure line (UMFS, NEH, Clinton) controls in a replicated field trial conducted at six grow-out sites in ME, MA, RI, CT, and NJ. The growth and survival of each replicate was monitored at each site every two months. Sampling continued until the oysters reached approximately 18 months of age. We estimated the line-specific survival, whole wet weight, meat weight and yield for each replicate. Oysters in each line at each site were also monitored for the prevalence and intensity of the diseases MSX, SSO, Dermo and ROD using a variety of methods. Conditioning and spawning - Fifty oysters from the NEH, Clinton, UMFS and UMFS x F 1 hybrid lines were obtained and placed in quarantine at the Darling Marine Center Shellfish Hatchery in February of The oyster broods were acclimated to C and pulse-fed a mixed diet of Isochrysis galbana, Tetraselmis sp., Pavlova lutheri, and Chaetoceros muelleri. Oysters were strip-spawned on May 11, 2011; strip-spawning allowed us to equalize egg and sperm concentrations among parents and lines, to maximize the number of parents contributing to each line, and enabled us to spawn animals from each of the four groups in synchrony so that outcrosses could be condstructed and the eggs fertilized within a reasonable amount of time. For the strip spawning procedure, broodstock were shucked and immediately sexed by microscopically examining a small biopsy of gonad tissue at 40X magnification. Gonads were then scored with a disposable scalpel (changed after each oyster to prevent cross-contamination) and eggs were teased out of the gonad into 1 L cups containing filtered (1 μm), UV sterilized seawater (FSW) at C. Egg suspensions were washed through a 150 pm sieve to eliminate debris and retained on a 20 pm sieve where they were briefly rinsed with FSW and combined into 1 L beakers. Sperm was collected dry in a microcentrifuge tube and stored on ice until all animals were stripped. A dilute sperm suspension was made from each male oyster by combining a small aliquot of dry concentrated sperm with 1 ml of FSW in a microcentrifuge tube; the dilutions were made and kept on ice for no more than 30 min before being used. The eggs from UMFS females were mixed gently and an equal amount placed into individual 100 ml cups. Diluted sperm from each male was added to the individual cups at an approximate ratio of 100 2

10 sperm per egg, so that each cup contained the eggs from a single female and a single male. The eggs in each cup were examined microscopically after 1 h to check for development and more sperm was added if the proportion of developing eggs was less than 70%. After initial development, the resulting embryos were then combined into a single 20 L bucket containing FSW at C and allowed to continue development overnight until they were at the trochophore stage. We used the same approach with the Clinton and UMFS x NEH F 1 broods to produce offspring for the Clinton parental and UMFS x NEH F2 lines, respectively. The following morning trochophores from these lines were stocked into duplicate 350 L tanks at densities of eggs ml"1. Densities in the larval tanks were reduced to 10 larvae ml-1at the time o f first drain-down the following day (i.e., 2 d post-fertilization). We produced the UMFS x Clinton F 1 hybrid line by combining the sperm from UMFS males with the eggs from Clinton females and sperm from Clinton males with eggs from the UMFS females. The embryos produced from the Clinton and UMFS eggs were stocked into a separate 350 L tanks until the first drain down, at which time the early D-stage larvae were combined and used to stock duplicate 350 L tanks at a density of 10 larvae ml-1. The broodstock from the NEH line partially spawned on their own on May 6, 2011, prior to our being able to strip-spawn them. We quickly dropped the water temperature in the tank holding these broods and obtained additional NEH broods from the Marine Biological Laboratory. We attempted to stripspawn both sets of NEH broods on May 11th. Although we recovered adequate volumes of viable sperm from 25 male NEH oysters and in excess of 3 million eggs from NEH females, the survival of NEH embryos was poor, suggesting poor egg quality. Similarly, the survival of F1 hybrids produced by crossing eggs from NEH females with sperm from UMFS males was low. We discarded the embryos from these crosses after the first few days of larval culture. Thus, the UMFS x NEH F1 in this project is composed of offspring from matings between UMFS females and NEH males, only. We purchased triploid NEH (NEH 3N) seed from Mook Sea Farms, a commercial hatchery located in Maine, to serve as the NEH line control for this project. This complication is unfortunate, but provided us an opportunity to compare the performance of NEH 3N oysters to our other experimental lines at sites throughout the Northeast. Triploid NEH have been quite popular among growers in the region due to their reported fast growth and disease resistance in southern New England (Guo et al. 2008). In addition, UMFS broodstock used in this project differs from those used in our previous efforts. The broods that we obtained for this line from the Pemaquid Oyster Co. were sampled from among the survivors of an outbreak of MSX that occurred on the Damariscotta River, ME in While generally being considered resistant to only ROD, these broods may also have obtained resistance to MSX. The genetic status and disease resistance characteristics of the six lines deployed in our experiment are indicated in Table 1. Table 1. The disease resistance characteristics of the six genetic lines of Eastern oysters included in this field trial. Line Genetic Status Disease Resistance UMFS Pure Line ROD. MSX? Clinton R u e Line ROD. MSX NEH (3N) Pure Line (Triploid) MSX. Dermo UMFS x NEH (2N) F 1 Hybrid Unknown UMFS x NEH (2N) F2 Hybrid Unknown UMFS x Clinton F 1 Hybrid Unknown 3

11 Larval, post-set, and nursery culture - Larvae were raised at 24 C at a salinity of ppt and fed a mixed diet of Isochrysis galbana, Pavlova lutheri, Pavlova sp. (CCMP459), and Thalassiosira pseudonana (strain 3H). Tanks were drained down every other day. Larvae were gradually thinned to density of 1-2 individuals ml-1ml prior to setting but size culling was kept to a minimum. Two tanks of each line were maintained to serve as a back-up in case of culture failure in one of the tanks. Upon becoming competent, larvae were placed into screened (180 μm) floating wooden tray culture units with microcultch. These bins were rinsed daily and tanks were cleaned and refilled with FSW every other day. Post-set animals were pulse-fed a mixed diet of live algae. Bins were washed every four days and graded as needed while maintaining the smallest healthy grade. When the animals reached a size of 1-2 mm they were placed in a recirculating upweller tank. Every 1-2 d the tank water was replaced with FSW and the animals were pulse-fed a mixture of live algae (species listed previously in addition to Tetraselmis chui, Chaetocerous meulerii, and Rhodomonas sp.) and algae paste (Reed Mariculture, Campbell, CA, Shellfish diet 1800). While oysters of this size are normally put into a system where they feed off natural phytoplankton, our protocol of maintaining the seed in a filtered sea-water system and feeding cultured algae was included as an added safeguard against potential transfer of infectious agents from raw Damariscotta River water to grow-out sites used in this experiment. A health certification was performed for each oyster line by I. Sunila using histology and Ray s Thioglycollate Medium before transplantations. No infectious parasites or diseases were observed. Upon reaching a size of 2-3 mm, oyster seed from each line was placed inside three replicate windowscreen mesh inserts at a density of 1000 seed per insert. These mesh inserts were placed inside the typical ADPI bag system used for surface culture of oysters at each site, or held in nursery upweller systems at each site. Replicates were identified by inclusion of a color and number-coded cattle tag in the bag (NASCO; Modesto, CA). Three replicate inserts from each of the six lines were deployed at the seven sites given in Table 2 (below). Overall, we deployed and monitored the performance of 126,000 oyster seed in our field trial (7 sites x 6 lines x oysters each). Field Monitoring - Table 2. Typical disease pressure observed at six sites at which we deployed eastern The performance oysters in this field trial. Due to logistical issues, oysters deployed at a seventh site (mortality and (Edgartown. MA) were lost d u ring the project. size) of oysters in all deployed bags Site/Location Symbol Grower Type of Site was assessed Cape Shore. New Jersey NJ Guo Dermo Clinton. Connecticut CT Sunila every 2-3 months Dermo. MSX. SSO. ROD Westerly, Rhode Island during two RI Gardner Dermo. M SX SSO Wellfleet. Massachusetts MA Cummings Dermo. SSO growing seasons Damariscotta River. Maine ME1 Davis MSX. SSO. ROD (2011 and 2012). Bagaduce River. Maine ME2 Leach Disease free The general measuring protocol consisted of recording the total wet-packed volume, haphazard sampling of individual animals and measuring the shell height (distance from the umbo to outer growing edge) of 50 or more animals while separating out mortalities. Animals that were cemented together ( doubles ) were assumed to be growing abnormally and excluded from the sample. The number of mortalities from the sample was tallied and general comments on the seed batch noted. Occasionally, it was necessary to cull contaminating organisms (e.g., mussels) or dead shell, if mortalities were significant. Bags were thinned by reducing the oyster volume by half when necessary with the thinned oysters 4

12 going to the cooperating grower s crop. Final grow-out densities were approximately oysters per bag depending on the site. Oysters were overwintered at each grow-out site; we made no attempt to standardize overwintering procedure. Oysters were redeployed in the spring o f 2012 following the schedule typically adhered to by each grower with the exception of the oysters at the Damariscotta River site which were overwintered in the Darling Marine Center Hatchery. Due to a miscommunication with our industry partners at the Edgartown Pond site, all oysters were lost at the end of the first grow-out season (2011) reducing our field trial to six sites. At the conclusion of the field trial in the late fall of 2012, oysters in all replicates were harvested, and final measurements of survival (counts of live and dead oysters), shell height, shell width, and wet weight were taken and yield for each replicate bag estimated as the number of kilograms harvested per 1000 oysters deployed (replicate yield = (replicate survival)* (replicate mean wet weight)). For each of these measurements, we compared line performance at each site by separate one-way analyses of variance (ANOVA) using the General Linear Model option in SYSTAT (ver. 12), with line as a fixed factor; the raw data were transformed to correct for heteroscedasticity among residuals, if necessary. To test for line by site interactions for each measurement variable, we combined the data from all sites and employed two-way ANOVA in which line and site were fixed-factors and included the line by site interaction term. As with the single factor ANOVAs, the raw data were transformed to correct for heteroscedasticity, if necessary. One concern for a broodstock program focused primarily on fast growth is that the resulting lines may have enhanced shell growth, but poor quality meat or shape which, in turn, could result in an eventual loss in market value. To assess shape we compared the shell height (SH) to width (SW), and shell height to shell inflation (SI) ratios (see Figure 1). This approach simultaneously investigates the relative width and cupping of the shell. Since deeply cupped oysters usually contain more meat, they are more desirable in the market place while a line that is losing its cup would be undesirable. We also determined the condition index (Crosby and Gale, 1990) by measuring individual live weights, sacrificing the oyster and separating all tissue from the shell and then determining the shell weight, and the dry meat weight. Condition index typically compares the dry meat weight of an oyster to the internal cavity volume. However, Lawrence and Scott (1982) have pointed out that oyster meats have a density o f approximately 1 g cm-3, and internal cavity volume can be estimated by subtracting the weight o f the shell valves in air from the total live weight of an oyster in air. Thus, condition index for at least 30 oysters from the each line at each site was estimated as: Figure 1. Snell measurements used to estimate shape and degree of cupping (as per Harding,2007). CI = dry tissue weight (g) x 1000 / [total live weight (g) - shell valve weight (g)] The shell height, dry weight, condition index, and the shell dimension ratios were statistically analyzed using nested analysis of variance (ANOVA) with oyster line as the main effect and replicate bag within line as a nested effect using the General Linear Model option in SYSTAT. 5

13 Disease Testing: Roseovarius Oyster Disease (ROD) - In September, October and November of 2011, 30 oysters from each line were randomly sampled from among the three replicates held at the Damariscotta River site. An additional 30 oysters from each line were sampled from the Clinton, CT, site in October of Upon sampling, these oysters were frozen and shipped to the University of Maine, Orono and processed for ROD diagnosis. Subsequently, the oysters were removed from the freezer and allowed to thaw at room temperature (~23 C) for 1 h. Upon thawing, the valves of each oyster were separated and all soft tissue removed and stored in 95% ethanol for later genetic testing. The shell of each oyster was examined for evidence of ROD symptoms. The degree of shell curvature for the top valve was determined using a qualitative scale. Valves were scored as normal (N) if the top and bottom valves both extended to the greatest shell height, mild (M) if there was slight overgrowth of the top valve, and severe (S) if there was extreme overgrowth of the top valve. In a similar manner, the shells were examined for evidence of excess conchiolin. Valves were scored as normal (N) if they had a white, smooth interior, discolored (D) if there was some yellowish or brownish coloration on the interior of the shell, mild (M) if there were minor conchiolin deposits on the interior, or severe (S) if there was a solid ring of conchiolin present. Following the methods of Ford and Borrero (2001), we combined the degree of curvature and conchiolin deposition measures to generate a numerical score of ROD symptoms for each oyster (Table 3). The statistical significance of line-specific differences in the severity of ROD symptoms was determined using a chi-squared test of independence (i.e., R x C contingency analysis). Oysters < 25 mm shell height are most susceptible to ROD (Davis and Barber 1999). Thus, there is the possibility that line-specific differences in growth could result in differences in susceptibility to ROD and the severity of ROD symptoms. After removing oyster tissue, we measured the height, width and inflation of the cupped valve of each oyster. The statistical significance of differences in mean shell height, width and inflation were examined using one-way analysis of variance (ANOVA). Finally, we compared line-specific differences in the severity of symptoms to line-specific patterns of mortality estimated during our regular field monitoring. We had intended to also use a QPCR-based assay based on the methods described by Maloy et al. (2005) to detect ROD. However, the reliability of this assay has not been verified and further development of the assay was beyond the scope of our project. Table 3. Scoring system for determining the severity of ROD symptoms in oysters sampled from the Damariscotta River and Clinton sites. We combined the degree of curvature and conchiolin deposition to generate a numerical score of ROD symptoms as per Ford and Borrero (2001). ROD Score Conchiolin Shell Curvature 0 None Normal 1 2 Discoloration on inner surface of shell Dark discoloration and light conchiolin deposits 3 Heavy conchiolin ring Slight Curvature to both valves Valves curved and margins do not meet. Severe overgrowth of top valve and severe curvature Disease Testing: MSX, SSO, Dermo and Histopathological Changes - In October of 2012, tissues were collected from 30 oysters for each line at each location (total n=1260) for estimating the level of MSX, SSO, ROD and Dermo infection in each line. Oysters were processed for histological examination using standard methods (Howard et al. 2004). A sample size of 30 allows detection of parasites with 10% prevalence with a 95% confidence limit (Post and Millest 1991). A 4-mm cross-section was excised from each oyster and fixed in Davidson s fixative. The samples were then dehydrated and 6

14 embedded into paraffin. Paraffin blocks were processed by Massachusetts Histology Services to produce a 5μm Hematoxylin-eosin stained slide. Slides were diagnosed for infectious agents, such as lesions caused by viruses, bacteria, prokaryotes (Chlamydia, Mycoplams and Rickettsia), ciliates, gregarines, and trematodes, and haplosporidian parasites MSX (H. nelsoni) and SSO (H. costale). Haplosporidian infections were staged as initial, intermediate, advanced, and terminal to allow prognosis of survival time. In previous NRAC work, we found that commensal species typically considered harmless to oysters caused pathological lesions (i.e., became parasitic) in some oyster lines introduced to new areas suggesting that such organisms act as opportunistic pathogens to oysters when the oysters are grown under sub-optimal conditions. Slides were thus diagnosed for inflammatory responses (acute and chronic), degenerations (inclusions, vacuolization, ceroidosis), cell and tissue death (necrosis, apoptosis), growth derangements (hyperplasia, metaplasia), hemodynamic and fluid derangements (edema, hemorrhage), and neoplasia (benign or malignant) according to Sunila (1998) to assess not only pathological lesions caused by parasites, but also the general health of the oysters. The prevalence of pathogens and lesions were compared using standard statistical methods between different strains and different sites. Disease Testing: Quantitative PCR Detection o f M SX and Dermo - In our initial proposal, we planned to conduct real-time quantitive polymerase chain reaction (qpcr)-based assays to quantify the level of P. marinus, R. crassostreae and H. nelsoni infection in the same oysters from our histopathological investigations (above). Dr. Marta Gomez-Chairri (University of Rhode Island) has been working to develop an assay that simultaneously detects MSX and Dermo in the same qpcr reaction. In addition to the use of improved primer sets for MSX and Dermo, this assay also allows for the detection of SSO. Given the potential benefits of this improved assay, we opted to work with Dr. Gomez-Chiarri instead of using the protocols outlined in our original proposal and have sent tissue samples to her lab from the oysters that were part of our standard histological examinations. Refinement of the protocol and cross-referencing of PCR and histopathological results is on-going. At this point in time, we do not have sufficient results from the qpcr analysis to include in this report. Objective 2 - In order for our broodstock development program to be successful it is critical that the lines we produce are available to industry. Our hatcheries at the Darling Marine Center and Marine Biological Laboratory were established as research hatcheries and do not sell seed to commercial growers. If the lines we are testing are to be adopted by the industry, it will be important for commercial hatcheries to become multiplier sites for larger scale production and distribution of oyster seed. There is currently no mechanism in place to ensure smooth and reliable distribution of pathogen-free stocks to industry. As our development program moves forward, regular communication with both hatcheries and growers will be critical to the adoption and equitable distribution of improved lines while honoring any intellectual property rights agreements (IPRAs), like those that govern the distribution of the NEH line. To date, the bulk of our communication has been with participating growers and through presentations at regional and national meetings. We plan to expand our efforts through additional communications, including 1) dissemination of this final report along with a brief summary of the results from both this project and its predecessor (NRAC #Q /Q ) via and posting on the website of the East Coast Shellfish Growers Association (ECSGA), 2) production of NRAC Fact Sheets summarizing the current knowledge of disease-resistance and production characteristics of extant lines and their hybrids, and on the history of the lines, and 3) formal journal publications on the production and disease-resistant characteristics of the lines used in our project. In addition, we are planning to convene a meeting of interested growers and hatchery operators at the NACE 2015 meeting in Portland, to review project results, the availability o f lines and 7

15 future research plans. In the meantime, the DMC hatchery is propagating the Clinton and UMFS lines while the Haskins Shellfish lab continues to maintain the NEH lines and make these available to hatcheries and growers in the region. R e s u l t s a n d D is c u s s io n Cumulative Mortality - There was substantial variation in the degree and timing of line-specific mortality at each of the six field sites in this grow-out trial. At Cape Shore, NJ, Clinton, CT, and Westerly, RI, mortality increased steeply and peaked during the summer and early fall of the second season. For example, at the Cape Shore sites, cumulative mortality was less than 30% for all six lines up through June of 2012, but then increased rapidly and exceeded 70% by September for most lines (Fig. 2, top left panel). Total cumulative mortality for the UMFS, Clinton and UMFS x Clinton F1 was nearly 90%, while NEH was significantly lower at 42%. Mortality for UMFS x NEH hybrids (both the F1 and F2 lines) was intermediate to these two extremes. We observed similar patterns in mortality at Clinton, CT site where cumulative mortality for the NEH line was ten-fold lower than the UMFS, Clinton and UMFS x Clinton lines and six-fold lower than the UMFS x NEH hybrid lines (Fig. 2, top right panel). At Westerly RI site, mortality also increased during the second season, but the increase in mortality was not nearly as steep as it was at the Cape Shore and Clinton sites (Fig. 2, middle left panel). Final cumulative mortality for the Clinton line (60%) exceeded that observed for the other five lines and was nearly double that observed for the NEH line (30%) which had the lowest mortality at the Westerly site. Mortality for the UMFS, UMFS x Clinton F 1, and UMFS x NEH F2 lines ranged from 40-50% while the line with the lowest mortality at the Westerly site was the UMFS x NEH F 1 hybrid line (20%). Mortality at the Wellfleet, MA site was the lowest in our field trial. Although we observed increased mortality in the Clinton line in the fall of 2012, final cumulative mortality was less than 20% for all lines at this site (Fig. 2, middle right panel). At the two sites in Maine, the differences in line-specific mortality were evident much earlier than at the sites in southern New England, typically occurring during the first few months following deployment. At both the Damariscotta River and Bagaduce River sites, mortality in the NEH line exceeded 50% by November (dashed box in lower panels of Fig. 2). Cumulative mortality for the Clinton line also increased dramatically over this same time period, although it peaked at a much smaller value (~25%). In contrast, cumulative mortality during the fall of 2011 for the UMFS and UMFS x NEH and UMFS x Clinton hybrid lines was less than 20% at the two Maine sites. The variation in mortality during the first fall of our field trial at the Damariscotta River site was highly correlated with the line-specific appearance of symptoms of ROD (see Disease Pressure discussion, below). These differences in mortality persisted to the end of our field trial; at the conclusion of the field trial, final cumulative mortality for NEH line exceeded 75% at both the Damariscotta River and Bagaduce River sites and was significantly higher than the cumulative mortality for the other five lines in our study. Final cumulative mortality for the Clinton line also exceeded 30% at both Maine sites while cumulative mortality for the UMFS, and hybrid lines typically was less than 20%. (Fig. 3) Patterns o f Growth Variation - Variation in growth among sites was substantial (Figs. 4 and 5) and estimates of the average shell height and average wet weight in each replicate at each site were highly correlated (r = 0.865; Fig. 6). The average shell height for all six lines exceeded 75mm, a size at which oysters are typically sent to market, at the Bagaduce River, ME, Damariscotta River, ME, Wellfleet, MA, and Westerly, RI sites (Fig. 4). At the Clinton, CT site only the NEH 3N oysters reached market 8

16 Date Figure 2. Variation in line-specific cumulative mortality over two year grow-out period for three genetic lines and their and interline hybrids deployed in triplicate bags at six culture sites in the Northeast. Oyster seed was deployed in the late summer of 2011 and the field trial was completed in November of 2012 when oysters at most sites had reached market size. The dashed boxes indicate periods of high mortality; the correlation between these mortality events and disease pressure at each sites are discussed in the text. 9

17 Figure 3. Final cumulative mortality for three parental lines of eastern oysters and their interline hybrids deployed at six sites (site designations are given in Table 2). Each bar represents the mean total mortality (±s.e.) for three replicate bags of each line. Line-specific means at each location that are significantly different at P<0.05 are indicated by letters above each bar. size, while the average size for the other lines (69-73mm) was slightly below 75mm. None of the lines reached market size at the Cape Shore site. Significant differences in oyster growth were detected only at southern New England sites. At the Wellfleet site, ANOVA detected a significant difference in shell height and wet weight. Despite a larger average shell height and wet weight for the NEH triploids at this site, Bonferroni post-hoc analyses of multiple pairwise comparisons (at an experiment-wide error rate of 0.05) failed to detect significant differences among individual lines. At the same time, the NEH 3N line outgrew the other lines at the Westerly, Clinton and Cape Shore sites. At Cape Shore and Westerly, the final shell height for the UMFS x NEH F 1 hybrid line was 4 to 10 mm smaller, respectively, than the average shell height of the NEH 3N line, but these differences were not statistically significant. Overall, poor growth for the UMFS and Clinton parental lines, and UMFS x Clinton F1 and UMFS x NEH F2 hybrid lines at the Cape Shore and Clinton sites was reflected in significantly lower shell heights and wet weights when compared to the size of the NEH 3N oysters. In contrast, there was no statistically significant difference in either shell height (Fig. 4) or wet weight (Fig. 5) among lines at the two Maine sites. The high degree of variation in line-specific growth among sites resulted in a highly significant interaction term in the two ANOVA for both shell height and wet weight (Table 4).

18 Figure 4. Mean shell height for three parental (Clinton, UMFS, and NEH) and three hybrid (UM x CL, UM x NEH F 1 and UM x NEH F2) lines of eastern oysters cultured at six sites in the Northeast. Bars represent the mean of three replicate means ± one standard error and letters above each bar indicate means at each site that differ from one another in pair-wise comparisons (P<0.05) while an asterisk indicates site where the ANOVA was significant (P<0.05) but individual pair-wise comparisons were not. Figure 5. Mean wet weight for three parental (Clinton, UMFS, and NEH) and three hybrid (UM x CL, UM x NEH F 1 and UM x NEH F2) lines of eastern oysters cultured at six sites in the Northeast. Bars represent the mean of three replicate means ± one standard error and letters above each bar indicate means at each site that differ from one another in pair-wise comparisons (P<0.05) while an asterisk indicates site where the ANOVA was significant (P<0.05) but individual pair-wise comparisons were not..

19 Table 4. Site by line interactions for Mortality. Shell Height, Wet Weight, and Yield. Due to heteroscedasticity caused by uneven variance among data from separate sites, the latter three variables were log transformed prior to analysis. Variable R2 Interaction SS df E rro r SS df F-ratio p-value Mortality <0.001 Log(Shell Height) <0.01 Log(Wet Weight) <0.01 Log(Yield) <0.001 Figure 6. Relationship between final wet weight and shell height for oysters in replicate bags of all six genetic lines cultured at six industry sites in this field trial. 12

20 Patterns in Yield - Significant differences in yield, measured as the mass (kg) of oysters recovered from the initial planting of 1000 seed in each replicate, were observed among lines at each of the sites where oysters were deployed (Fig. 7). Given both high survival and very fast growth, the highest yields were observed at the Wellfleet site. Although we did not detect line-specific differences in shell height or wet weight at this site, the combined effects of growth and survival generated significant linespecific differences in yield; the NEH 3N had 20% higher yield compared to the next best line (UMFS x NEH F 1) and 36 % higher than the Clinton line. The average yield at Westerly (38 kg) was only 50% of that observed at the Wellfleet site (79 kg), and the average yield at the two Maine sites was lower still (28-29 kg). Similar to our analysis on growth, we detected a significant site x line (genotype by environment) interaction for yield (Table 4). Overall, the yield in each replicate bag was highly correlated with both survival (inverse of cumulative mortality) and wet weight, as would be expected from the method by which we estimated yield (Fig. 8; Table 5). However, at the Westerly and Wellfleet sites, where there were only subtle differences in survival among lines, the correlation between yield and wet weight was higher than the correlation between yield and survival. In contrast, at the other four sites (Cape Shore, Clinton, Damariscotta River, and Bagaduce River), line-specific differences in yield were driven primarily by the differences in survival and growth had much less of an effect (Table 5). The specific disease pressures that are likely driving these differences in survival and yield are discussed below. Figure 7. Mean yield per 1000 oysters deployed for three parental (Clinton, UMFS, and NEH) and three hybrid (UM x CL, UM x NEH F 1 and UM x NEH F2) lines of eastern oysters cultured at six sites in the Northeast (site designations are given in Table 2). Bars represent the mean of three replicate means ± one standard error and letters above each bar indicate means at each site that differ from one another in pair-wise comparisons (P<0.05). 13